only let filesort() count rows for SQL_CALC_ROWS if it's using priority queue

Update InnoDB to 5.6.14
Apply MySQL-5.6 hack for MySQL Bug#16434374
Move Aria-only HA_RTREE_INDEX from my_base.h to maria_def.h (breaks an assert in InnoDB)
Fix InnoDB memory leak

Use Breaks: not Conflicts:

don't add entries to acl_roles_mappings hash for non-existing grantees.

recognize the context better:
always treat the barename as a username in the username context

use 'user_name' for the error message (fixed, after get_current_user()),
not the original tmp_user_name, as it comes from the parser (host == NULL).

set_role::light_check() was incorrect

MDEV-5221: User auto-creation does not work upon GRANT <role>

spider suites: #varname is a valid syntax for a variable name, don't use it for comments

- the problem was caused by EXPLAIN INSERT SELECT. For that statement, 
  the code would call select_insert::prepare2(), which would call 
  handler->ha_start_bulk_insert().  The corresponding handler->end_bulk_insert() 
  call is made from select_insert::send_eof or select_insert::abort_result_set 
  which are never called for EXPLAIN INSERT SELECT.
- Fixed by re-using approach of mysql-5.6: don't call ha_start_bulk_insert() if 
  we are in EXPLAIN.

Undo the previous band-aid fix in psergey@askmonty.org-20130802141209-4dqfvx2db8acxwbl.
Kristian has made a proper fix, which uses a different approach.

Fix for MDEV-4117 @@global.relay_log_purge not per-master, conflicts between different masters in multisource replication

The fix is to not change @relay_log_purge as part of the CHANGE MASTER.
(There is no logical reason why this is done in the current source)

mysql-test/suite/rpl/r/rpl_slave_status.result:
  Ensure that CHANGE MASTER doesn't change relay_log_purge
mysql-test/suite/rpl/t/rpl_slave_status.test:
  Ensure that CHANGE MASTER doesn't change relay_log_purge
sql/sql_repl.cc:
  Don't change relay_log_purge in CHANGE MASTER

Fix all suite.pm files that had errors and test files that were skipped because of that

* remove is_embedded.require

because some suites may be disabled unless a plugin is available,
and compiled-in plugins are only known after collect_mysqld_features().

Additionally move mysql_install_db() after everything else it's collected -
it's a slow process, better do it after all the information is gathered
(and reported to the user).

plugin.cmake does it automatically for all plugins

* rename "xtradb" combination to be called "innodb"
* disable xtradb_plugin embedded tests (because of RECOMPILE_FOR_EMBEDDED)

(it was used only for plugins in mysqld)

support ha_innodb.so as a dynamic plugin.
* remove obsolete *,innodb_plugin.rdiff files
* s/--plugin-load=/--plugin-load-add=/
* MYSQL_PLUGIN_IMPORT glob_hostname[]
* use my_error instead of push_warning_printf(ER_DEFAULT)
* don't use tdc_size and tc_size in a module

update test cases (XtraDB is 5.6.14, InnoDB is 5.6.10)
* copy new tests over
* disable some tests for (old) InnoDB
* delete XtraDB tests that no longer apply

small compatibility changes:
* s/HTON_EXTENDED_KEYS/HTON_SUPPORTS_EXTENDED_KEYS/
* revert unnecessary InnoDB changes to make it a bit closer to the upstream

fix XtraDB to compile on Windows (both as a static and a dynamic plugin)

disable XtraDB on Windows (deadlocks) and where no atomic ops are available (e.g. CentOS 5)


storage/innobase/handler/ha_innodb.cc:
  revert few unnecessary changes to make it a bit closer to the original InnoDB
storage/innobase/include/univ.i:
  correct the version to match what it was merged from

modified:
  storage/connect/odbconn.cpp